idnits 2.17.1
draft-ietf-sfc-nsh-00.txt:
Checking boilerplate required by RFC 5378 and the IETF Trust (see
https://trustee.ietf.org/license-info):
----------------------------------------------------------------------------
No issues found here.
Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt:
----------------------------------------------------------------------------
No issues found here.
Checking nits according to https://www.ietf.org/id-info/checklist :
----------------------------------------------------------------------------
== There are 3 instances of lines with non-RFC6890-compliant IPv4 addresses
in the document. If these are example addresses, they should be changed.
== There are 16 instances of lines with private range IPv4 addresses in the
document. If these are generic example addresses, they should be changed
to use any of the ranges defined in RFC 6890 (or successor): 192.0.2.x,
198.51.100.x or 203.0.113.x.
Miscellaneous warnings:
----------------------------------------------------------------------------
== The copyright year in the IETF Trust and authors Copyright Line does not
match the current year
-- The document date (March 24, 2015) is 3312 days in the past. Is this
intentional?
Checking references for intended status: Proposed Standard
----------------------------------------------------------------------------
(See RFCs 3967 and 4897 for information about using normative references
to lower-maturity documents in RFCs)
== Unused Reference: 'RFC0791' is defined on line 1320, but no explicit
reference was found in the text
== Unused Reference: 'RFC2784' is defined on line 1328, but no explicit
reference was found in the text
== Unused Reference: 'VXLAN-gpe' is defined on line 1350, but no explicit
reference was found in the text
-- Obsolete informational reference (is this intentional?): RFC 5226
(Obsoleted by RFC 8126)
Summary: 0 errors (**), 0 flaws (~~), 6 warnings (==), 2 comments (--).
Run idnits with the --verbose option for more detailed information about
the items above.
--------------------------------------------------------------------------------
2 Network Working Group P. Quinn, Ed.
3 Internet-Draft Cisco Systems, Inc.
4 Intended status: Standards Track U. Elzur, Ed.
5 Expires: September 25, 2015 Intel
6 March 24, 2015
8 Network Service Header
9 draft-ietf-sfc-nsh-00.txt
11 Abstract
13 This draft describes a Network Service Header (NSH) inserted onto
14 encapsulated packets or frames to realize service function paths.
15 NSH also provides a mechanism for metadata exchange along the
16 instantiated service path.
18 1. Requirements Language
20 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
21 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
22 document are to be interpreted as described in RFC 2119 [RFC2119].
24 Status of this Memo
26 This Internet-Draft is submitted in full conformance with the
27 provisions of BCP 78 and BCP 79.
29 Internet-Drafts are working documents of the Internet Engineering
30 Task Force (IETF). Note that other groups may also distribute
31 working documents as Internet-Drafts. The list of current Internet-
32 Drafts is at http://datatracker.ietf.org/drafts/current/.
34 Internet-Drafts are draft documents valid for a maximum of six months
35 and may be updated, replaced, or obsoleted by other documents at any
36 time. It is inappropriate to use Internet-Drafts as reference
37 material or to cite them other than as "work in progress."
39 This Internet-Draft will expire on September 25, 2015.
41 Copyright Notice
43 Copyright (c) 2015 IETF Trust and the persons identified as the
44 document authors. All rights reserved.
46 This document is subject to BCP 78 and the IETF Trust's Legal
47 Provisions Relating to IETF Documents
48 (http://trustee.ietf.org/license-info) in effect on the date of
49 publication of this document. Please review these documents
50 carefully, as they describe your rights and restrictions with respect
51 to this document. Code Components extracted from this document must
52 include Simplified BSD License text as described in Section 4.e of
53 the Trust Legal Provisions and are provided without warranty as
54 described in the Simplified BSD License.
56 Table of Contents
58 1. Requirements Language . . . . . . . . . . . . . . . . . . . . 2
59 2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 4
60 2.1. Definition of Terms . . . . . . . . . . . . . . . . . . . 4
61 2.2. Problem Space . . . . . . . . . . . . . . . . . . . . . . 6
62 3. Network Service Header . . . . . . . . . . . . . . . . . . . . 8
63 3.1. Network Service Header Format . . . . . . . . . . . . . . 8
64 3.2. NSH Base Header . . . . . . . . . . . . . . . . . . . . . 8
65 3.3. Service Path Header . . . . . . . . . . . . . . . . . . . 10
66 3.4. NSH MD-type 1 . . . . . . . . . . . . . . . . . . . . . . 10
67 3.4.1. Mandatory Context Header Allocation Guidelines . . . . 11
68 3.5. NSH MD-type 2 . . . . . . . . . . . . . . . . . . . . . . 12
69 3.5.1. Optional Variable Length Metadata . . . . . . . . . . 13
70 4. NSH Actions . . . . . . . . . . . . . . . . . . . . . . . . . 15
71 5. NSH Encapsulation . . . . . . . . . . . . . . . . . . . . . . 17
72 6. NSH Usage . . . . . . . . . . . . . . . . . . . . . . . . . . 18
73 7. NSH Proxy Nodes . . . . . . . . . . . . . . . . . . . . . . . 19
74 8. Fragmentation Considerations . . . . . . . . . . . . . . . . . 20
75 9. Service Path Forwarding with NSH . . . . . . . . . . . . . . . 21
76 9.1. SFFs and Overlay Selection . . . . . . . . . . . . . . . . 21
77 9.2. Mapping NSH to Network Overlay . . . . . . . . . . . . . . 23
78 9.3. Service Plane Visibility . . . . . . . . . . . . . . . . . 24
79 9.4. Service Graphs . . . . . . . . . . . . . . . . . . . . . . 24
80 10. Policy Enforcement with NSH . . . . . . . . . . . . . . . . . 26
81 10.1. NSH Metadata and Policy Enforcement . . . . . . . . . . . 26
82 10.2. Updating/Augmenting Metadata . . . . . . . . . . . . . . . 27
83 10.3. Service Path ID and Metadata . . . . . . . . . . . . . . . 29
84 11. NSH Encapsulation Examples . . . . . . . . . . . . . . . . . . 30
85 11.1. GRE + NSH . . . . . . . . . . . . . . . . . . . . . . . . 30
86 11.2. VXLAN-gpe + NSH . . . . . . . . . . . . . . . . . . . . . 30
87 11.3. Ethernet + NSH . . . . . . . . . . . . . . . . . . . . . . 31
88 12. Security Considerations . . . . . . . . . . . . . . . . . . . 32
89 13. Open Items for WG Discussion . . . . . . . . . . . . . . . . . 33
90 14. Contributors . . . . . . . . . . . . . . . . . . . . . . . . . 34
91 15. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 37
92 16. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 38
93 16.1. NSH EtherType . . . . . . . . . . . . . . . . . . . . . . 38
94 16.2. Network Service Header (NSH) Parameters . . . . . . . . . 38
95 16.2.1. NSH Base Header Reserved Bits . . . . . . . . . . . . 38
96 16.2.2. MD Type Registry . . . . . . . . . . . . . . . . . . . 38
97 16.2.3. TLV Class Registry . . . . . . . . . . . . . . . . . . 39
98 16.2.4. NSH Base Header Next Protocol . . . . . . . . . . . . 39
99 17. References . . . . . . . . . . . . . . . . . . . . . . . . . . 40
100 17.1. Normative References . . . . . . . . . . . . . . . . . . . 40
101 17.2. Informative References . . . . . . . . . . . . . . . . . . 40
102 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 42
104 2. Introduction
106 Service functions are widely deployed and essential in many networks.
107 These service functions provide a range of features such as security,
108 WAN acceleration, and server load balancing. Service functions may
109 be instantiated at different points in the network infrastructure
110 such as the wide area network, data center, campus, and so forth.
112 The current service function deployment models are relatively static,
113 and bound to topology for insertion and policy selection.
114 Furthermore, they do not adapt well to elastic service environments
115 enabled by virtualization.
117 New data center network and cloud architectures require more flexible
118 service function deployment models. Additionally, the transition to
119 virtual platforms requires an agile service insertion model that
120 supports elastic service delivery; the movement of service functions
121 and application workloads in the network and the ability to easily
122 bind service policy to granular information such as per-subscriber
123 state are necessary.
125 The approach taken by NSH is composed of the following elements:
127 1. Service path identification
129 2. Transport independent per-packet/frame service metadata.
131 3. Optional variable TLV metadata.
133 NSH is designed to be easy to implement across a range of devices,
134 both physical and virtual, including hardware platforms.
136 An NSH aware control plane is outside the scope of this document.
138 The SFC Architecture document [SFC-arch] provides an overview of a
139 service chaining architecture that clearly defines the roles of the
140 various elements and the scope of a service function chaining
141 encapsulation.
143 2.1. Definition of Terms
145 Classification: Locally instantiated policy and customer/network/
146 service profile matching of traffic flows for identification of
147 appropriate outbound forwarding actions.
149 SFC Network Forwarder (NF): SFC network forwarders provide network
150 connectivity for service functions forwarders and service
151 functions. SFC network forwarders participate in the network
152 overlay used for service function chaining as well as in the SFC
153 encapsulation.
155 Service Function Forwarder (SFF): A service function forwarder is
156 responsible for delivering traffic received from the NF to one or
157 more connected service functions, and from service functions to
158 the NF.
160 Service Function (SF): A function that is responsible for specific
161 treatment of received packets. A service function can act at the
162 network layer or other OSI layers. A service function can be a
163 virtual instance or be embedded in a physical network element.
164 One of multiple service functions can be embedded in the same
165 network element. Multiple instances of the service function can
166 be enabled in the same administrative domain.
168 Service Node (SN): Physical or virtual element that hosts one or
169 more service functions and has one or more network locators
170 associated with it for reachability and service delivery.
172 Service Function Chain (SFC): A service function chain defines an
173 ordered set of service functions that must be applied to packets
174 and/or frames selected as a result of classification. The implied
175 order may not be a linear progression as the architecture allows
176 for nodes that copy to more than one branch. The term service
177 chain is often used as shorthand for service function chain.
179 Service Function Path (SFP): The instantiation of a SFC in the
180 network. Packets follow a service function path from a classifier
181 through the requisite service functions
183 Network Node/Element: Device that forwards packets or frames based
184 on outer header information. In most cases is not aware of the
185 presence of NSH.
187 Network Overlay: Logical network built on top of existing network
188 (the underlay). Packets are encapsulated or tunneled to create
189 the overlay network topology.
191 Network Service Header: Data plane header added to frames/packets.
192 The header contains information required for service chaining, as
193 well as metadata added and consumed by network nodes and service
194 elements.
196 Service Classifier: Function that performs classification and
197 imposes an NSH. Creates a service path. Non-initial (i.e.
198 subsequent) classification can occur as needed and can alter, or
199 create a new service path.
201 Service Hop: NSH aware node, akin to an IP hop but in the service
202 overlay.
204 Service Path Segment: A segment of a service path overlay.
206 NSH Proxy: Acts as a gateway: removes and inserts NSH on behalf of a
207 service function that is not NSH aware.
209 2.2. Problem Space
211 Network Service Header (NSH) addresses several limitations associated
212 with service function deployments today.
214 1. Topological Dependencies: Network service deployments are often
215 coupled to network topology. Such dependency imposes constraints
216 on the service delivery, potentially inhibiting the network
217 operator from optimally utilizing service resources, and reduces
218 the flexibility. This limits scale, capacity, and redundancy
219 across network resources.
221 2. Service Chain Construction: Service function chains today are
222 most typically built through manual configuration processes.
223 These are slow and error prone. With the advent of newer service
224 deployment models the control/management planes provide not only
225 connectivity state, but will also be increasingly utilized for
226 the creation of network services. Such a control/management
227 planes could be centralized, or be distributed.
229 3. Application of Service Policy: Service functions rely on topology
230 information such as VLANs or packet (re) classification to
231 determine service policy selection, i.e. the service function
232 specific action taken. Topology information is increasingly less
233 viable due to scaling, tenancy and complexity reasons. The
234 topological information is often stale, providing the operator
235 with inaccurate placement that can result in suboptimal resource
236 utilization. Furthermore topology-centric information often does
237 not convey adequate information to the service functions, forcing
238 functions to individually perform more granular classification.
240 4. Per-Service (re)Classification: Classification occurs at each
241 service function independent from previously applied service
242 functions. More importantly, the classification functionality
243 often differs per service function and service functions may not
244 leverage the results from other service functions.
246 5. Common Header Format: Various proprietary methods are used to
247 share metadata and create service paths. An open header provides
248 a common format for all network and service devices.
250 6. Limited End-to-End Service Visibility: Troubleshooting service
251 related issues is a complex process that involve both network-
252 specific and service-specific expertise. This is especially the
253 case when service function chains span multiple DCs, or across
254 administrative boundaries. Furthermore, the physical and virtual
255 environments (network and service) can be highly divergent in
256 terms of topology and that topological variance adds to these
257 challenges.
259 7. Transport Dependence: Service functions can and will be deployed
260 in networks with a range of transports requiring service
261 functions to support and participate in many transports (and
262 associated control planes) or for a transport gateway function to
263 be present.
265 Please see the Service Function Chaining Problem Statement [SFC-PS]
266 for a more detailed analysis of service function deployment problem
267 areas.
269 3. Network Service Header
271 A Network Service Header (NSH) contains metadata and service path
272 information that are added to a packet or frame and used to create a
273 service plane. The packets and the NSH are then encapsulated in an
274 outer header for transport.
276 The service header is added by a service classification function - a
277 device or application - that determines which packets require
278 servicing, and correspondingly which service path to follow to apply
279 the appropriate service.
281 3.1. Network Service Header Format
283 An NSH is composed of a 4-byte base header, a 4-byte service path
284 header and context headers, as shown in Figure 1 below.
286 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
287 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
288 | Base Header |
289 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
290 | Service Path Header |
291 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
292 | |
293 ~ Context Headers ~
294 | |
295 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
297 Figure 1: Network Service Header
299 Base header: provides information about the service header and the
300 payload protocol.
302 Service Path Header: provide path identification and location within
303 a path.
305 Context headers: carry opaque metadata and variable length encoded
306 information.
308 3.2. NSH Base Header
310 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
311 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
312 |Ver|O|C|R|R|R|R|R|R| Length | MD Type | Next Protocol |
313 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
314 Figure 2: NSH Base Header
316 Base Header Field Descriptions
318 Version: The version field is used to ensure backward compatibility
319 going forward with future NSH updates.
321 O bit: Indicates that this packet is an operations and management
322 (OAM) packet. SFF and SFs nodes MUST examine the payload and take
323 appropriate action (e.g. return status information).
325 OAM message specifics and handling details are outside the scope of
326 this document.
328 C bit: Indicates that a critical metadata TLV is present (see Section
329 3.4.2). This bit acts as an indication for hardware implementers to
330 decide how to handle the presence of a critical TLV without
331 necessarily needing to parse all TLVs present. The C bit MUST be set
332 to 1 if one or more critical TLVs are present.
334 All other flag fields are reserved.
336 Length: total length, in 4-byte words, of the NSH header, including
337 optional variable TLVs.
339 MD Type: indicates the format of NSH beyond the base header and the
340 type of metadata being carried. This typing is used to describe the
341 use for the metadata. A new registry will be requested from IANA for
342 the MD Type.
344 NSH defines two MD types:
346 0x1 which indicates that the format of the header includes fixed
347 length context headers.
349 0x2 which does not mandate any headers beyond the base header and
350 service path header, and may contain optional variable length context
351 information.
353 The format of the base header is invariant, and not described by MD
354 Type.
356 NSH implementations MUST support MD-Type 0x1, and SHOULD support MD-
357 Type 0x2.
359 Next Protocol: indicates the protocol type of the original packet. A
360 new IANA registry will be created for protocol type.
362 This draft defines the following Next Protocol values:
364 0x1 : IPv4
365 0x2 : IPv6
366 0x3 : Ethernet
368 3.3. Service Path Header
370 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
371 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
372 | Service Path ID | Service Index |
373 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
375 Service path ID (SPI): 24 bits
376 Service index (SI): 8 bits
378 Figure 3: NSH Service Path Header
380 Service Path Identifier (SPI): identifies a service path.
381 Participating nodes MUST use this identifier for path selection. An
382 administrator can use the service path value for reporting and
383 troubleshooting packets along a specific path.
385 Service Index (SI): provides location within the service path.
386 Service index MUST be decremented by service functions or proxy nodes
387 after performing required services. MAY be used in conjunction with
388 service path for path selection. Service Index is also valuable when
389 troubleshooting/reporting service paths. In addition to location
390 within a path, SI can be used for loop detection.
392 3.4. NSH MD-type 1
394 When the base header specifies MD Type 1, NSH defines four 4-byte
395 mandatory context headers, as per Figure 4. These headers must be
396 present and the format is opaque as depicted in Figure 5.
398 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
399 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
400 |Ver|O|C|R|R|R|R|R|R| Length | MD-type=0x1 | Next Protocol |
401 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
402 | Service Path ID | Service Index |
403 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
404 | Mandatory Context Header |
405 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
406 | Mandatory Context Header |
407 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
408 | Mandatory Context Header |
409 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
410 | Mandatory Context Header |
411 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
413 Figure 4: NSH MD-type=0x1
415 0 1 2 3
416 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
417 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
418 | Context data |
419 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
421 Figure 5: Context Header
423 3.4.1. Mandatory Context Header Allocation Guidelines
425 0 1 2 3
426 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
427 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
428 | Network Platform Context |
429 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
430 | Network Shared Context |
431 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
432 | Service Platform Context |
433 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
434 | Service Shared Context |
435 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
436 Figure 6: Context Data Significance
438 Figure 6, above, and the following examples of context header
439 allocation are guidelines that illustrate how various forms of
440 information can be carried and exchanged via NSH.
442 Network platform context: provides platform-specific metadata shared
443 between network nodes. Examples include (but are not limited to)
444 ingress port information, forwarding context and encapsulation type.
446 Network shared context: metadata relevant to any network node such as
447 the result of edge classification. For example, application
448 information, identity information or tenancy information can be
449 shared using this context header.
451 Service platform context: provides service platform specific metadata
452 shared between service functions. This context header is analogous
453 to the network platform context, enabling service platforms to
454 exchange platform-centric information such as an identifier used for
455 load balancing decisions.
457 Service shared context: metadata relevant to, and shared, between
458 service functions. As with the shared network context,
459 classification information such as application type can be conveyed
460 using this context.
462 The data center[dcalloc] and mobility[moballoc] context header
463 allocation drafts provide guidelines for the semantics of NSH fixed
464 context headers in each respective environment.
466 3.5. NSH MD-type 2
468 When the base header specifies MD Type 2, NSH defines variable length
469 only context headers. There may be zero or more of these headers as
470 per the length field.
472 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
473 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
474 |Ver|O|C|R|R|R|R|R|R| Length | MD-type=0x2 | Next Protocol |
475 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
476 | Service Path ID | Service Index |
477 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
478 | |
479 ~ Optional Variable Length Context Headers ~
480 | |
481 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
482 Figure 7: NSH MD-type=0x2
484 3.5.1. Optional Variable Length Metadata
486 NSH MD Type 2 MAY contain optional variable length context headers.
487 The format of these headers is as described below.
489 0 1 2 3
490 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
491 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
492 | TLV Class | Type |R|R|R| Len |
493 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
494 | Variable Metadata |
495 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
497 Figure 8: Variable Context Headers
499 TLV Class: describes the scope of the "Type" field. In some cases,
500 the TLV Class will identify a specific vendor, in others, the TLV
501 Class will identify specific standards body allocated types.
503 Type: the specific type of information being carried, within the
504 scope of a given TLV Class. Value allocation is the responsibility
505 of the TLV Class owner.
507 The most significant bit of the Type field indicates whether the TLV
508 is mandatory for the receiver to understand/process. This
509 effectively allocates Type values 0 to 127 for non-critical options
510 and Type values 128 to 255 for critical options. Figure 7 below
511 illustrates the placement of the Critical bit within the Type field.
513 +-+-+-+-+-+-+-+-+
514 |C| Type |
515 +-+-+-+-+-+-+-+-+
517 Figure 9: Critical Bit Placement Within the TLV Type Field
519 Encoding the criticality of the TLV within the Type field is
520 consistent with IPv6 option types.
522 If a receiver receives an encapsulated packet containing a TLV with
523 the Critical bit set in the Type field and it does not understand how
524 to process the Type, it MUST drop the packet. Transit devices MUST
525 NOT drop packets based on the setting of this bit.
527 Reserved bits: three reserved bit are present for future use. The
528 reserved bits MUST be zero.
530 Length: Length of the variable metadata, in 4-byte words.
532 4. NSH Actions
534 Service header aware nodes - service classifiers, SFF, SF and NSH
535 proxies, have several possible header related actions:
537 1. Insert or remove service header: These actions can occur at the
538 start and end respectively of a service path. Packets are
539 classified, and if determined to require servicing, a service
540 header imposed. The last node in a service path, an SFF, removes
541 the NSH. A service classifier MUST insert an NSH. At the end of
542 a service function chain, the last node operating on the service
543 header MUST remove it.
545 A service function can re-classify data as required and that re-
546 classification might result in a new service path. In this case,
547 the SF acts as a logical classifier as well. When the logical
548 classifier performs re-classification that results in a change of
549 service path, it MUST remove the existing NSH and MUST impose a
550 new NSH with the base header reflecting the new path.
552 2. Select service path: The base header provides service chain
553 information and is used by SFFs to determine correct service path
554 selection. SFFs MUST use the base header for selecting the next
555 service in the service path.
557 3. Update a service header: NSH aware service functions MUST
558 decrement the service index. A service index = 0 indicates that
559 a packet MUST be dropped by the SFF performing NSH-based
560 forwarding.
562 Service functions MAY update context headers if new/updated
563 context is available.
565 If an NSH proxy (see Section 7) is in use (acting on behalf of a
566 non-NSH-aware service function for NSH actions), then the proxy
567 MUST update service index and MAY update contexts. When an NSH
568 proxy receives an NSH-encapsulated packet, it removes the NSH
569 before forwarding it to an NSH unaware SF. When it receives a
570 packet back from an NSH unaware SF, it re-encapsulates it with
571 the NSH, decrementing the service index.
573 4. Service policy selection: Service function instances derive
574 policy selection from the service header. Context shared in the
575 service header can provide a range of service-relevant
576 information such as traffic classification. Service functions
577 SHOULD use NSH to select local service policy.
579 Figure 10 maps each of the four actions above to the components in
580 the SFC architecture that can perform it.
582 +----------------+--------------------+-------+---------------+-------+
583 | | Insert or remove |Select | Update a |Service|
584 | | service header |service|service header |Policy |
585 | +------+------+------+ path +---------------+Select-|
586 | |Insert|Remove|Remove| | Dec. |Update |ion |
587 | | | | and | |Service|Context| |
588 | Component | | |Insert| | Index |Header | |
589 +----------------+------+------+------+-------+-------+-------+-------+
590 |Service Classif-| + | | | | | + | |
591 |ication Function| | | | | | | |
592 + -------------- + ---- + ---- + ---- + ----- + ----- + ----- + ----- +
593 |Service Function| | + | | + | | + | |
594 |Forwarder(SFF) | | | | | | | |
595 + -------------- + ---- + ---- + ---- + ----- + ----- + ----- + ----- +
596 |Service | | | | | + | + | + |
597 |Function (SF) | | | | | | | |
598 + -------------- + ---- + ---- + ---- + ----- + ----- + ----- + ----- +
599 |NSH Proxy | + | + | | | + | + | |
600 +----------------+------+------+------+-------+-------+-------+-------+
602 Figure 10: NSH Action and Role Mapping
604 5. NSH Encapsulation
606 Once NSH is added to a packet, an outer encapsulation is used to
607 forward the original packet and the associated metadata to the start
608 of a service chain. The encapsulation serves two purposes:
610 1. Creates a topologically independent services plane. Packets are
611 forwarded to the required services without changing the
612 underlying network topology.
614 2. Transit network nodes simply forward the encapsulated packets as
615 is.
617 The service header is independent of the encapsulation used and is
618 encapsulated in existing transports. The presence of NSH is
619 indicated via protocol type or other indicator in the outer
620 encapsulation.
622 See Section 11 for NSH encapsulation examples.
624 6. NSH Usage
626 The NSH creates a dedicated service plane, that addresses many of the
627 limitations highlighted in Section 2.2. More specifically, NSH
628 enables:
630 1. Topological Independence: Service forwarding occurs within the
631 service plane, via a network overlay, the underlying network
632 topology does not require modification. Service functions have
633 one or more network locators (e.g. IP address) to receive/send
634 data within the service plane, the NSH contains an identifier
635 that is used to uniquely identify a service path and the services
636 within that path.
638 2. Service Chaining: NSH contains path identification information
639 needed to realize a service path. Furthermore, NSH provides the
640 ability to monitor and troubleshoot a service chain, end-to-end
641 via service-specific OAM messages. The NSH fields can be used by
642 administrators (via, for example a traffic analyzer) to verify
643 (account, ensure correct chaining, provide reports, etc.) the
644 path specifics of packets being forwarded along a service path.
646 3. Metadata Sharing: NSH provides a mechanism to carry shared
647 metadata between network devices and service function, and
648 between service functions. The semantics of the shared metadata
649 is communicated via a control plane to participating nodes.
650 Examples of metadata include classification information used for
651 policy enforcement and network context for forwarding post
652 service delivery.
654 4. Transport Agnostic: NSH is transport independent and is carried
655 in an overlay, over existing underlays. If an existing overlay
656 topology provides the required service path connectivity, that
657 existing overlay may be used.
659 7. NSH Proxy Nodes
661 In order to support NSH-unaware service functions, an NSH proxy is
662 used. The proxy node removes the NSH header and delivers the
663 original packet/frame via a local attachment circuit to the service
664 function. Examples of a local attachment circuit include, but are
665 not limited to: VLANs, IP in IP, GRE, VXLAN. When complete, the
666 service function returns the packet to the NSH proxy via the same or
667 different attachment circuit.
669 NSH is re-imposed on packets returned to the proxy from the non-NSH-
670 aware service.
672 Typically, an SFF will act as an NSH-proxy when required.
674 An NSH proxy MUST perform NSH actions as described in Section 4.
676 8. Fragmentation Considerations
678 Work in progress
680 9. Service Path Forwarding with NSH
682 9.1. SFFs and Overlay Selection
684 As described above, NSH contains a service path identifier (SPI) and
685 a service index (SI). The SPI is, as per its name, an identifier.
686 The SPI alone cannot be used to forward packets along a service path.
687 Rather the SPI provide a level of indirection between the service
688 path/topology and the network transport. Furthermore, there is no
689 requirement, or expectation of an SPI being bound to a pre-determined
690 or static network path.
692 The service index provides an indication of location within a service
693 path. The combination of SPI and SI provides the identification and
694 location of a logical SF (locator and order). The logical SF may be
695 a single SF, or a set of SFs that are equivalent. In the latter
696 case, the SFF provides load distribution amongst the collection of
697 SFs as needed. SI may also serve as a mechanism for loop detection
698 with in a service path since each SF in the path decrements the
699 index; an index of 0 indicates that a loop occurred and packet must
700 be discarded.
702 This indirection -- path ID to overlay -- creates a true service
703 plane. That is the SFF/SF topology is constructed without impacting
704 the network topology but more importantly service plane only
705 participants (i.e. most SFs) need not be part of the network overlay
706 topology and its associated infrastructure (e.g. control plane,
707 routing tables, etc.). As mentioned above, an existing overlay
708 topology may be used provided it offers the requisite connectivity.
710 The mapping of SPI to transport occurs on an SFF. The SFF consults
711 the SPI/ID values to determine the appropriate overlay transport
712 protocol (several may be used within a given network) and next hop
713 for the requisite SF. Figure 10 below depicts an SPI/SI to network
714 overlay mapping.
716 +-------------------------------------------------------+
717 | SPI | SI | NH | Transport |
718 +-------------------------------------------------------+
719 | 10 | 3 | 1.1.1.1 | VXLAN-gpe |
720 | 10 | 2 | 2.2.2.2 | nvGRE |
721 | 245 | 12 | 192.168.45.3 | VXLAN-gpe |
722 | 10 | 9 | 10.1.2.3 | GRE |
723 | 40 | 9 | 10.1.2.3 | GRE |
724 | 50 | 7 | 01:23:45:67:89:ab | Ethernet |
725 | 15 | 1 | Null (end of path) | None |
726 +-------------------------------------------------------+
727 Figure 11: SFF NSH Mapping Example
729 Additionally, further indirection is possible: the resolution of the
730 required SF function locator may be a localized resolution on an
731 SFF,rather than a service function chain control plane
732 responsibility, as per figures 11 and 12 below.
734 +-------------------+
735 | SPI | SI | NH |
736 +-------------------+
737 | 10 | 3 | SF2 |
738 | 245 | 12 | SF34 |
739 | 40 | 9 | SF9 |
740 +-------------------+
742 Figure 12: NSH to SF Mapping Example
744 +-----------------------------------+
745 | SF | NH | Transport |
746 +-----------------------------------|
747 | SF2 | 10.1.1.1 | VXLAN-gpe |
748 | SF34| 192.168.1.1 | UDP |
749 | SF9 | 1.1.1.1 | GRE |
750 +-----------------------------------+
752 Figure 13: SF Locator Mapping Example
754 Since the SPI is a representation of the service path, the lookup may
755 return more than one possible next-hop within a service path for a
756 given SF, essentially a series of weighted (equally or otherwise)
757 overlay links to be used (for load distribution, redundancy or
758 policy), see Figure 13. The metric depicted in Figure 13 is an
759 example to help illustrated weighing SFs. In a real network, the
760 metric will range from a simple preference (similar to routing next-
761 hop), to a true dynamic composite metric based on some service
762 function-centric state (including load, sessions sate, capacity,
763 etc.)
764 +----------------------------------+
765 | SPI | SI | NH | Metric |
766 +----------------------------------+
767 | 10 | 3 | 10.1.1.1 | 1 |
768 | | | 10.1.1.2 | 1 |
769 | | | | |
770 | 20 | 12 | 192.168.1.1 | 1 |
771 | | | 10.2.2.2 | 1 |
772 | | | | |
773 | 30 | 7 | 10.2.2.3 | 10 |
774 | | | 10.3.3.3 | 5 |
775 +----------------------------------+
776 (encap type omitted for formatting)
778 Figure 14: NSH Weighted Service Path
780 9.2. Mapping NSH to Network Overlay
782 As described above, the mapping of SPI to network topology may result
783 in a single overlay path, or it might result in a more complex
784 topology. Furthermore, the SPIx to overlay mapping occurs at each
785 SFF independently. Any combination of topology selection is
786 possible.
788 Examples of mapping for a topology:
790 1. Next SF is located at SFFb with locator 10.1.1.1
791 SFFa mapping: SPI=10 --> VXLAN-gpe, dst-ip: 10.1.1.1
793 2. Next SF is located at SFFc with multiple locator for load
794 distribution purposes:
795 SFFb mapping: SPI=10 --> VXLAN-gpe, dst_ip:10.2.2.1, 10.2.2.2,
796 10.2.2.3, equal cost
798 3. Next SF is located at SFFd with two path to SFFc, one for
799 redundancy:
800 SFFc mapping: SPI=10 --> VXLAN-gpe, dst_ip:10.1.1.1 cost=10,
801 10.1.1.2, cost=20
803 In the above example, each SFF makes an independent decision about
804 the network overlay path and policy for that path. In other words,
805 there is no a priori mandate about how to forward packets in the
806 network (only the order of services that must be traversed).
808 The network operator retains the ability to engineer the overlay
809 paths as required. For example, the overlay path between service
810 functions forwarders may utilize traffic engineering, QoS marking, or
811 ECMP, without requiring complex configuration and network protocol
812 support to be extended to the service path explicitly. In other
813 words, the network operates as expected, and evolves as required, as
814 does the service function plane.
816 9.3. Service Plane Visibility
818 The SPI and SI serve an important function for visibility into the
819 service topology. An operator can determine what service path a
820 packet is "on", and its location within that path simply by viewing
821 the NSH information (packet capture, IPFIX, etc.). The information
822 can be used for service scheduling and placement decisions,
823 troubleshooting and compliance verification.
825 9.4. Service Graphs
827 In some cases, a service path is exactly that -- a linear list of
828 service functions that must be traversed. However, increasingly, the
829 "path" is actually a true directed graph. Furthermore, within a
830 given service topology several directed graphs may exist with packets
831 moving between graphs based on non-initial classification (usually
832 performed by a service function). Note: strictly speaking a path is
833 a form of graph; the intent is to distinguish between a directed
834 graph and a path.
836 ,---. ,---. ,---.
837 / \ / \ / \
838 ( SF2 ) ( SF7 ) ( SF3 )
839 ,------\ +. \ / \ /
840 ; |---' `-. `---'\ `-+-'
841 | : : \ ;
842 | \ | : ;
843 ,-+-. `. ,+--. : |
844 / \ '---+ \ \ ;
845 ( SF1 ) ( SF6 ) \ /
846 \ / \ +--. : /
847 `---' `---' `-. ,-+-. /
848 `+ +'
849 ( SF4 )
850 \ /
851 `---'
853 Figure 15: Service Graph Example
855 The SPI/SI combination provides a simple representation of a directed
856 graph, the SPI represents a graph ID; and the SI a node ID. The
857 service topology formed by SPI/SI support cycles, weighting, and
858 alternate topology selection, all within the service plane. The
859 realization of the network topology occurs as described above: SPI/ID
860 mapping to an appropriate transport and associated next network hops.
862 NSH-aware services receive the entire header, including the SPI/SI.
863 An SF can now, based on local policy, alter the SPI, which in turn
864 effects both the service graph, and in turn the selection of overlay
865 at the SFF. The figure below depicts the policy associated with the
866 graph in Figure 14 above. Note: this illustrates multiple graphs and
867 their representation; it does not depict the use of metadata within a
868 single service function graph.
870 +---------------------------------------------------------------------+
871 | SPI: 21 Bob: SF7 |
872 | SPI: 20 Bad : SF2 --> SF6 --> SF4 |
873 |SPI: 10 SF1 --> SF2 --> SF6 SPI: 22 Alice: SF3 |
874 | SPI: 30 Good: SF4 |
875 | SPI:31 Bob: SF7 |
876 | SPI:32 Alice: SF3 |
877 +---------------------------------------------------------------------+
879 Figure 16: Service Graphs Using SPI
881 This example above does not show the mapping of the service topology
882 to the network overlay topology. As discussed in the sections above,
883 the overlay selection occurs as per network policy.
885 10. Policy Enforcement with NSH
887 10.1. NSH Metadata and Policy Enforcement
889 As described in Section 3, NSH provides the ability to carry metadata
890 along a service path. This metadata may be derived from several
891 sources, common examples include:
893 Network nodes: Information provided by network nodes can indicate
894 network-centric information (such as VRF or tenant) that may be
895 used by service functions, or conveyed to another network node
896 post-service pathing.
898 External (to the network) systems: External systems, such as
899 orchestration systems, often contain information that is valuable
900 for service function policy decisions. In most cases, this
901 information cannot be deduced by network nodes. For example, a
902 cloud orchestration platform placing workloads "knows" what
903 application is being instantiated and can communicate this
904 information to all NSH nodes via metadata.
906 Service functions: Service functions often perform very detailed
907 and valuable classification. In some cases they may terminate,
908 and be able to inspect encrypted traffic. SFs may update, alter
909 or impose metadata information.
911 Regardless of the source, metadata reflects the "result" of
912 classification. The granularity of classification may vary. For
913 example, a network switch might only be able to classify based on a
914 5-tuple, whereas, a service function may be able to inspect
915 application information. Regardless of granularity, the
916 classification information can be represented in NSH.
918 Once the data is added to NSH, it is carried along the service path,
919 NSH-aware SFs receive the metadata, and can use that metadata for
920 local decisions and policy enforcement. The following two examples
921 highlight the relationship between metadata and policy:
923 +-------------------------------------------------+
924 | ,---. ,---. ,---. |
925 | / \ / \ / \ |
926 | ( SCL )-------->( SF1 )--------->( SF2 ) |
927 | \ / \ / \ / |
928 | `---' `---' `---' |
929 |5-tuple: Permit Inspect |
930 |Tenant A Tenant A AppY |
931 |AppY |
932 +-------------------------------------------------+
934 Figure 17: Metadata and Policy
936 +-------------------------------------------------+
937 | ,---. ,---. ,---. |
938 | / \ / \ / \ |
939 | ( SCL )-------->( SF1 )--------->( SF2 ) |
940 | \ / \ / \ / |
941 | `-+-' `---' `---' |
942 | | Permit Deny AppZ |
943 | +---+---+ employees |
944 | | | |
945 | +-------+ |
946 | external |
947 | system: |
948 | Employee |
949 | App Z |
950 +-------------------------------------------------+
952 Figure 18: External Metadata and Policy
954 In both of the examples above, the service functions perform policy
955 decisions based on the result of the initial classification: the SFs
956 did not need to perform re-classification, rather they relied on a
957 antecedent classification for local policy enforcement.
959 10.2. Updating/Augmenting Metadata
961 Post-initial metadata imposition (typically performed during initial
962 service path determination), metadata may be augmented or updated:
964 1. Metadata Augmentation: Information may be added to NSH's existing
965 metadata, as depicted in Figure 18. For example, if the initial
966 classification returns the tenant information, a secondary
967 classification (perhaps a DPI or SLB) may augment the tenant
968 classification with application information. The tenant
969 classification is still valid and present, but additional
970 information has been added to it.
972 2. Metadata Update: Subsequent classifiers may update the initial
973 classification if it is determined to be incorrect or not
974 descriptive enough. For example, the initial classifier adds
975 metadata that describes the trafic as "internet" but a security
976 service function determines that the traffic is really "attack".
977 Figure 19 illustrates an example of updating metadata.
979 +-------------------------------------------------+
980 | ,---. ,---. ,---. |
981 | / \ / \ / \ |
982 | ( SCL )-------->( SF1 )--------->( SF2 ) |
983 | \ / \ / \ / |
984 | `-+-' `---' `---' |
985 | | Inspect Deny |
986 | +---+---+ employees employee+ |
987 | | | Class=AppZ appZ |
988 | +-------+ |
989 | external |
990 | system: |
991 | Employee |
992 | |
993 +-------------------------------------------------+
995 Figure 19: Metadata Augmentation
997 +-------------------------------------------------+
998 | ,---. ,---. ,---. |
999 | / \ / \ / \ |
1000 | ( SCL )-------->( SF1 )--------->( SF2 ) |
1001 | \ / \ / \ / |
1002 | `---' `---' `---' |
1003 |5-tuple: Inspect Deny |
1004 |Tenant A Tenant A attack |
1005 | --> attack |
1006 +-------------------------------------------------+
1008 Figure 20: Metadata Update
1010 10.3. Service Path ID and Metadata
1012 Metadata information may influence the service path selection since
1013 the service path identifier can represent the result of
1014 classification. A given SPI can represent all or some of the
1015 metadata, and be updated based on metadata classification results.
1016 This relationship provides the ability to create a dynamic services
1017 plane based on complex classification without requiring each node to
1018 be capable of such classification, or requiring a coupling to the
1019 network topology. This yields service graph functionality as
1020 described in Section 9.4. Figure 20 illustrates an example of this
1021 behavior.
1023 +----------------------------------------------------+
1024 | ,---. ,---. ,---. |
1025 | / \ / \ / \ |
1026 | ( SCL )-------->( SF1 )--------->( SF2 ) |
1027 | \ / \ / \ / |
1028 | `---' `---' \ `---' |
1029 |5-tuple: Inspect \ Original |
1030 |Tenant A Tenant A \ next SF |
1031 | --> DoS \ |
1032 | \ |
1033 | ,---. |
1034 | / \ |
1035 | ( SF10 ) |
1036 | \ / |
1037 | `---' |
1038 | DoS |
1039 | "Scrubber" |
1040 +----------------------------------------------------+
1042 Figure 21: Path ID and Metadata
1044 Specific algorithms for mapping metadata to an SPI are outside the
1045 scope of this draft.
1047 11. NSH Encapsulation Examples
1049 11.1. GRE + NSH
1051 IPv4 Packet:
1052 +----------+--------------------+--------------------+
1053 |L2 header | L3 header, proto=47|GRE header,PT=0x894F|
1054 +----------+--------------------+--------------------+
1055 --------------+----------------+
1056 NSH, NP=0x1 |original packet |
1057 --------------+----------------+
1059 L2 Frame:
1060 +----------+--------------------+--------------------+
1061 |L2 header | L3 header, proto=47|GRE header,PT=0x894F|
1062 +----------+--------------------+--------------------+
1063 ---------------+---------------+
1064 NSH, NP=0x3 |original frame |
1065 ---------------+---------------+
1067 Figure 22: GRE + NSH
1069 11.2. VXLAN-gpe + NSH
1071 IPv4 Packet:
1072 +----------+------------------------+---------------------+
1073 |L2 header | IP + UDP dst port=4790 |VXLAN-gpe NP=0x4(NSH)|
1074 +----------+------------------------+---------------------+
1075 --------------+----------------+
1076 NSH, NP=0x1 |original packet |
1077 --------------+----------------+
1079 L2 Frame:
1080 +----------+------------------------+---------------------+
1081 |L2 header | IP + UDP dst port=4790 |VXLAN-gpe NP=0x4(NSH)|
1082 +----------+------------------------+---------------------+
1083 ---------------+---------------+
1084 NSH,NP=0x3 |original frame |
1085 ---------------+---------------+
1087 Figure 23: VXLAN-gpe + NSH
1089 11.3. Ethernet + NSH
1091 IPv4 Packet:
1092 +-------------------------------+---------------+--------------------+
1093 |Outer Ethernet, ET=0x894F | NSH, NP = 0x1 | original IP Packet |
1094 +-------------------------------+---------------+--------------------+
1096 L2 Frame:
1097 +-------------------------------+---------------+----------------+
1098 |Outer Ethernet, ET=0x894F | NSH, NP = 0x3 | original frame |
1099 +-------------------------------+---------------+----------------+
1101 Figure 24: Ethernet + NSH
1103 12. Security Considerations
1105 As with many other protocols, NSH data can be spoofed or otherwise
1106 modified. In many deployments, NSH will be used in a controlled
1107 environment, with trusted devices (e.g. a data center) thus
1108 mitigating the risk of unauthorized header manipulation.
1110 NSH is always encapsulated in a transport protocol and therefore,
1111 when required, existing security protocols that provide authenticity
1112 (e.g. RFC 2119 [RFC6071]) can be used.
1114 Similarly if confidentiality is required, existing encryption
1115 protocols can be used in conjunction with encapsulated NSH.
1117 13. Open Items for WG Discussion
1119 1. MD type 1 metadata semantics specifics
1121 2. Bypass bit in NSH.
1123 3. Rendered Service Path ID (RSPID).
1125 14. Contributors
1127 This WG document originated as draft-quinn-sfc-nsh and had the
1128 following co-authors and contributors. The editors of this document
1129 would like to thank and recognize them and their contributions.
1130 These co-authors and contributors provided invaluable concepts and
1131 content for this document's creation.
1133 Surendra Kumar
1134 Cisco Systems
1135 smkumar@cisco.com
1137 Michael Smith
1138 Cisco Systems
1139 michsmit@cisco.com
1141 Jim Guichard
1142 Cisco Systems
1143 jguichar@cisco.com
1145 Rex Fernando
1146 Cisco Systems
1147 Email: rex@cisco.com
1149 Navindra Yadav
1150 Cisco Systems
1151 Email: nyadav@cisco.com
1153 Wim Henderickx
1154 Alcatel-Lucent
1155 wim.henderickx@alcatel-lucent.com
1157 Andrew Dolganow
1158 Alcaltel-Lucent
1159 Email: andrew.dolganow@alcatel-lucent.com
1161 Praveen Muley
1162 Alcaltel-Lucent
1163 Email: praveen.muley@alcatel-lucent.com
1165 Tom Nadeau
1166 Brocade
1167 tnadeau@lucidvision.com
1169 Puneet Agarwal
1170 puneet@acm.org
1172 Rajeev Manur
1173 Broadcom
1174 rmanur@broadcom.com
1176 Abhishek Chauhan
1177 Citrix
1178 Abhishek.Chauhan@citrix.com
1180 Joel Halpern
1181 Ericsson
1182 joel.halpern@ericsson.com
1184 Sumandra Majee
1185 F5
1186 S.Majee@f5.com
1188 David Melman
1189 Marvell
1190 davidme@marvell.com
1192 Pankaj Garg
1193 Microsoft
1194 Garg.Pankaj@microsoft.com
1196 Brad McConnell
1197 Rackspace
1198 bmcconne@rackspace.com
1200 Chris Wright
1201 Red Hat Inc.
1202 chrisw@redhat.com
1204 Kevin Glavin
1205 Riverbed
1206 kevin.glavin@riverbed.com
1208 Hong (Cathy) Zhang
1209 Huawei US R&D
1210 cathy.h.zhang@huawei.com
1212 Louis Fourie
1213 Huawei US R&D
1214 louis.fourie@huawei.com
1216 Ron Parker
1217 Affirmed Networks
1218 ron_parker@affirmednetworks.com
1220 Myo Zarny
1221 Goldman Sachs
1222 myo.zarny@gs.com
1224 15. Acknowledgments
1226 The authors would like to thank Nagaraj Bagepalli, Abhijit Patra,
1227 Peter Bosch, Darrel Lewis, Pritesh Kothari, Tal Mizrahi and Ken Gray
1228 for their detailed review, comments and contributions.
1230 A special thank you goes to David Ward and Tom Edsall for their
1231 guidance and feedback.
1233 Additionally the authors would like to thank Carlos Pignataro and
1234 Larry Kreeger for their invaluable ideas and contributions which are
1235 reflected throughout this draft.
1237 Lastly, Reinaldo Penno deserves a particular thank you for his
1238 architecture and implementation work that helped guide the protocol
1239 concepts and design.
1241 16. IANA Considerations
1243 16.1. NSH EtherType
1245 An IEEE EtherType, 0x894F, has been allocated for NSH.
1247 16.2. Network Service Header (NSH) Parameters
1249 IANA is requested to create a new "Network Service Header (NSH)
1250 Parameters" registry. The following sub-sections request new
1251 registries within the "Network Service Header (NSH) Parameters "
1252 registry.
1254 16.2.1. NSH Base Header Reserved Bits
1256 There are ten bits at the beginning of the NSH Base Header. New bits
1257 are assigned via Standards Action [RFC5226].
1259 Bits 0-1 - Version
1260 Bit 2 - OAM (O bit)
1261 Bits 2-9 - Reserved
1263 16.2.2. MD Type Registry
1265 IANA is requested to set up a registry of "MD Types". These are
1266 8-bit values. MD Type values 0, 1, 2, 254, and 255 are specified in
1267 this document. Registry entries are assigned by using the "IETF
1268 Review" policy defined in RFC 5226 [RFC5226].
1270 +---------+--------------+---------------+
1271 | MD Type | Description | Reference |
1272 +---------+--------------+---------------+
1273 | 0 | Reserved | This document |
1274 | | | |
1275 | 1 | NSH | This document |
1276 | | | |
1277 | 2 | NSH | This document |
1278 | | | |
1279 | 3..253 | Unassigned | |
1280 | | | |
1281 | 254 | Experiment 1 | This document |
1282 | | | |
1283 | 255 | Experiment 2 | This document |
1284 +---------+--------------+---------------+
1286 Table 1
1288 16.2.3. TLV Class Registry
1290 IANA is requested to set up a registry of "TLV Types". These are 16-
1291 bit values. Registry entries are assigned by using the "IETF Review"
1292 policy defined in RFC 5226 [RFC5226].
1294 16.2.4. NSH Base Header Next Protocol
1296 IANA is requested to set up a registry of "Next Protocol". These are
1297 8-bit values. Next Protocol values 0, 1, 2 and 3 are defined in this
1298 draft. New values are assigned via Standards Action [RFC5226].
1300 +---------------+-------------+---------------+
1301 | Next Protocol | Description | Reference |
1302 +---------------+-------------+---------------+
1303 | 0 | Reserved | This document |
1304 | | | |
1305 | 1 | IPv4 | This document |
1306 | | | |
1307 | 2 | IPv6 | This document |
1308 | | | |
1309 | 3 | Ethernet | This document |
1310 | | | |
1311 | 4..253 | Unassigned | |
1312 +---------------+-------------+---------------+
1314 Table 2
1316 17. References
1318 17.1. Normative References
1320 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791,
1321 September 1981.
1323 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
1324 Requirement Levels", BCP 14, RFC 2119, March 1997.
1326 17.2. Informative References
1328 [RFC2784] Farinacci, D., Li, T., Hanks, S., Meyer, D., and P.
1329 Traina, "Generic Routing Encapsulation (GRE)", RFC 2784,
1330 March 2000.
1332 [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an
1333 IANA Considerations Section in RFCs", BCP 26, RFC 5226,
1334 May 2008.
1336 [RFC6071] Frankel, S. and S. Krishnan, "IP Security (IPsec) and
1337 Internet Key Exchange (IKE) Document Roadmap", RFC 6071,
1338 February 2011.
1340 [SFC-PS] Quinn, P., Ed. and T. Nadeau, Ed., "Service Function
1341 Chaining Problem Statement", 2014, .
1345 [SFC-arch]
1346 Quinn, P., Ed. and J. Halpern, Ed., "Service Function
1347 Chaining (SFC) Architecture", 2014,
1348 .
1350 [VXLAN-gpe]
1351 Quinn, P., Agarwal, P., Kreeger, L., Lewis, D., Maino, F.,
1352 Yong, L., Xu, X., Elzur, U., and P. Garg, "Generic
1353 Protocol Extension for VXLAN",
1354 .
1356 [dcalloc] Guichard, J., Smith, M., and S. Kumar, "Network Service
1357 Header (NSH) Context Header Allocation (Data Center)",
1358 2014, .
1361 [moballoc]
1362 Napper, J. and S. Kumar, "NSH Context Header Allocation --
1363 Mobility", 2014, .
1366 Authors' Addresses
1368 Paul Quinn (editor)
1369 Cisco Systems, Inc.
1371 Email: paulq@cisco.com
1373 Uri Elzur (editor)
1374 Intel
1376 Email: uri.elzur@intel.com